基于免疫算法的故障诊断技术研究  

Fault diagnosis based on immune algorthm

作　　者：许宝杰[1] 徐小力[1] 吴国新[1] 谷玉海[1] 

机构地区：[1]北京信息科技大学机电工程学院,北京100192

出　　处：《北京信息科技大学学报（自然科学版）》2009年第2期24-28,共5页Journal of Beijing Information Science and Technology University

基　　金：北京市自然科学基金项目(3042006);北京市教育委员会科技项目(82052002)

摘　　要：针对故障先验知识和特征信息缺乏的情况,提出将免疫算法应用于机电系统的故障诊断的新方法。利用基于生物免疫学原理的负选择原理,给出了改进的免疫算法及其匹配规则和算法流程,建立了将改进的免疫故障诊断方法应用于机电系统的故障诊断体系。免疫算法实现故障诊断的过程分为二个阶段:免疫系统建立阶段和免疫功能执行阶段。在转子实验台上进行了原理性验证实验,进行实例验证。结果表明新方法可实现故障诊断基本功能。For the problem of lacking apriori knowledge of failure and characteristic information, a new method is proposed which applying immune algorithm to failure diagnosis of mechantronic systems. Using negative selection rule based on the principle of biology immunity, improved immune algorithm and its match rule and algorithm program are proposed. A new failure diagnosis structure applying improved immune diagnosing method for mechantronie systems is developed. The process of diagnosis by immune algorithm is divided to two steps：building of immune system and performing of immune system. Rotary machine failure diagnosis experiment is accomplished in principle. The result indicates the basic function of failure diagnosis is realized.

关 键 词：免疫算法 故障诊断 机电系统 状态监测 

分 类 号：TH17[机械工程—机械制造及自动化]